PET is a functional imaging technique in nuclear medicine that produces a three-dimensional image of functional processes in a living subject. Typically, a short-lived radioactive tracer isotope, such as fluorodeoxyglucose (FDG), is injected into the subject, where it becomes concentrated in a tissue of interest. As the radioisotope undergoes positron emission decay (also known as beta decay), it emits a positron (an antiparticle of the electron and having an opposite charge). The positron travels for a distance that is typically less than 1 mm (dependent on the type of isotope), during which time it loses kinetic energy. Once sufficiently decelerated it can interact with an electron. The electron and positron annihilate, which produces a pair of annihilation photons that move in approximately opposite directions. A significant fraction of the electron-positron annihilations results in two gamma rays or photons at 511 keV, emitted at approximately 180° apart and along a straight line of response (LOR) to pixelated gamma ray detectors that are positioned around the subject. The two gamma photons are detected within a certain coincidence timing window. With an appropriate time-resolution of detected pairs of gamma rays, the time-of-flight may be determined for each coincident event, which may be used to determine the location of the annihilation event. An image is generated based on the acquired gamma energy and hit-location data.
Statistics are obtained for many thousands of coincidence events. Typically, the coincidence events may be grouped into projection images, called sinograms. Known reconstruction techniques such as filtered back projection (FBP) may be used to reconstruct images from the projection images, resulting in a map that shows tissues in which the tracer has become concentrated, which can be interpreted by a physician or radiologist in the context of supporting a diagnosis or treatment for the subject.
PET is used for both medical and research applications. For instance, it may be used in clinical oncology to study tumors and search for metastases, or for clinical diagnosis of brain diseases. PET is also used to map brain and heart function, or to support drug development. PET is capable of detecting areas of molecular biology detail and may be used in a dedicated scanner.
Typically, such a scanner includes a patient table that provides for placement of the patient within the proper proximity of a bank of pixelated gamma ray detectors. Commonly, the bank of gamma ray detectors is positioned over 360° of the patient. The detectors convert incoming pairs of gamma rays to a light signal, which are then converted to electrical signals and processed in a data acquisition system. The processed data is passed to a computing device for imaging reconstruction. The amount of information obtained for imaging may be based on factors that include system noise. Noise may be generated, in one example, in which scatter occurs within the subject (where photons are deflected). In another example, noise may be generated based on random events in which two photons originating from two different annihilation events are incorrectly attributed to the same event—thereby incorrectly recorded as a coincidence pair arriving within the timing window.
A PET system may include a dedicated PET scanner, or may be combined with other known imaging modalities such as computed tomography (CT) or magnetic resonance imaging (MRI). In combined systems, both metabolic and anatomic information may be co-registered to provide combined information about both structure and biochemical activity.
Traditional PET detectors typically include arrays that are segmented scintillator pixels, each with a cross-section of approximately 10 mm2, in one example, coupled to photomultiplier tubes with couple centimeter diameters. Because of the mismatch in scintillator and photomultiplier sensitive areas, signal analysis methods such as “Anger logic” are commonly used to determine the gamma ray hit position. Anger logic is a procedure to obtain the position of incidence of a photon on the scintillator, which includes connecting photomultiplier outputs to a resistive network to obtain only four outputs. With these signals or outputs, the 2-dimensional position of the scintillation centroid is obtained using a simple geometric formula. However, the resulting spatial image typically shows large distortions and non-uniform energy resolution across the sensitive area.
Recent detector technology employs a photosensor called Silicon Photomultipliers (SiPM) that includes an array of avalanche photodiodes with total sensitive area matching the scintillator pixel. An avalanche photodiode (APD) is a highly sensitive semiconductor electronic device that exploits the photoelectric effect to convert light to electrical signal. A SiPM operates at high speeds and high gain by applying a reverse bias voltage (typically 100-200 V but in some cases as low as 26 to 30 V), and shows an internal current gain effect, a factor of one million in one example. The one-on-one SiPM-scintillator pixel coupling configuration provides excellent timing resolution and highly uniform energy resolution across the sensitive area. But because each pixel element convers a small area of a few mm2, a whole-body PET scanner based on this technology involves many tens of thousands of detector channels.
The electronic readout of the large number of SiPM detector channels is handled either by application specific ICs (ASICs) or by some type of electronic channel-reduction circuits. In an alternative, frontend electronics based on discrete components mounted on conventional printed circuit boards (PCBs) can be used. Currently the general approach is to use a resistive (or capacitive) network to derive weighted sums of the detector elements, which can be used to determine a hit position of the gamma ray. The drawback is that the timing and energy resolutions are degraded due to the large (summed) detector capacitance and noise. Another approach employs delay-line switched network architecture together with a field-programmable gate array (FPGA) to readout only the hit pixel. This improves timing and energy resolutions, but the resulting boards can be bulky, power hungry and expensive to build. Furthermore, existing readout technology is able to provide only 2-dimensional hit location information.
Thus, there is a need to reduce overall channel count in PET scanners and to provide 3-dimensional hit location to improve line-of-response determination.